wildcat_paris

What do you need to do with the XU4 "gleam ASAP"?

Quite expensive : Turris Omnia for gigabits/s fiber internet. for ADSL line, Lamobo-R1 is fine.

write protected? yes and no, that is the purpose of the sd_fusing.sh script to allow boot0 to be rewritten. https://github.com/igorpecovnik/lib/blob/a82dc77aee0c286de600192365922dc74741185f/patch/u-boot/u-boot-odroidxu-next/xu4_sd_fusing.patch 1/ the eMMC2SD converter Hardkernel provides cannot see boot0/1/etc. internal partitions. 2/ Only a XU4 booted from SD card can do that & use sd_fusing.sh script 3/ if you don't overwrite boot0 of eMMC, you have a creepy uboot 2012.07 which will probably fail with single partion/vanilla kernel. you cannot rely on Hardkernel for software, only Samsung Opensource Group does a great job.

Me. I haven't seen uboot 2012.07 working with a single partition. I can be wrong, of course. And anyway, without overwriting boot0 partition on eMMC with a proper uboot & blobs, I have tested, you are not gonna make it working. In the best case, SD could work. my next step would have been to test more fully how uboot 2016.09 & legacy kernel works (HDMI stuff & so on) but... read your words. A project is to attract people, not repeal them.

@zador.blood.stain @mrneilypops XU4 : uboot 2012.07 needs a dual boot partition. So uboot 2016.09 is needed as https://github.com/igorpecovnik/lib/pull/520 or a switch back to dual partitions for legacy kernel (with uboot 2016.09, a writing/booting to a SD card is needed, then somewhere in the SD card there is sd_fusing.sh to write the boot loader in "boot0" partition) @zador.blood.stain: better switch back to dual partition for legacy kernel on XU4 for the moment.

@zador.blood.stain You did good about the missing kernel options. it is running. (Still need the update to rename or use xu4.dtb instead of xu3) but ok!!!! http://pastebin.com/Gwyuh7Kv http://pastebin.com/sVjga3bL the diff http://pastebin.com/zyanC78b probably we need to unify boot.cmd now both legacy/vanilla can boot with single partition uboot 2016.09 edit: uboot 2016.09 & legacy kernel, the fan doesn't seem to work edit2: so both legacy/vanilla boots with uboot 2016.09 with the diff (on SD, eMMC to be confirmed)

@zador.blood.stain yes it "kind of" boots, but the system cannot go any further with root=<null> Sure the dtb file renaming is a "detail" I can handle.

Board Branch Boot Network HDMI Install Date Performed by link ----------------------------------------------------------------------------------------------------------------------------------------------------- Lamobo R1 Legacy ? ? ? ? ? ? Lamobo R1 Vanilla ? ? ? ? ? ? Odroid XU4 Legacy No - - SD 20161027 wildcat_paris http://pastebin.com/LA1sUa49 Odroid XU4 Vanilla Yes Yes console only SD 20161026 wildcat_paris ? comments: - I cannot find a proper/easy way to write a result table on the forum + impossible to use "tab" key to separate elements (only spaces) => is there an extension of the forum to add tables ??? I would propose to create a new "sticky" post: - first message is the summary of all reports for a beta version (I would propose I do the updates here) from... - testers can put their latest results in this thread and I would update the messages in the first post (as a summary) - maybe you may reinstall wildcat_paris with proper rights to manage the "testing thread" (& available to help more in forum as previously)

reading this uboot doesn't seem to be able to pass root= to the kernel edit: with uboot 2016.09 & legacy kernel I had to rename /boot/dtb/*xu4* with xu3

@zador.blood.stain I have tried uboot 2012.07 & boot.ini, uboot 2016.09 & boot.cmd (with and without UUID, removing ArmbianEnv.txt as well, putting /dev/mmcblk0p1 for root) but I end always with [ 7.905375] [c6] VFS: Cannot open root device "(null)" or unknown-block(0,0): error -6 [ 7.912127] [c6] Please append a correct "root=" boot option; here are the available partitions: [ 7.920918] [c6] b300 31276032 mmcblk0 driver: mmcblk [ 7.926612] [c6] b301 1214464 mmcblk0p1 65a81a70-01 [ 7.932308] [c6] 0800 488386576 sda driver: sd [ 7.937370] [c6] 0801 488383488 sda1 e72d222d-01 [ 7.942651] [c6] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)

@zador.blood.stain logs for legacy XU4 short: [ 7.905185] [c6] VFS: Cannot open root device "(null)" or unknown-block(0,0): error -6 [ 7.911928] [c6] Please append a correct "root=" boot option; here are the available partitions: [ 7.920718] [c6] b300 31276032 mmcblk0 driver: mmcblk [ 7.926411] [c6] b301 1214464 mmcblk0p1 01380d35-01 [ 7.932109] [c6] 0800 488386576 sda driver: sd [ 7.937171] [c6] 0801 488383488 sda1 e72d222d-01 [ 7.942452] [c6] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)

just an idea: would we create a subforum of Development for testing reports like Debian/Ubuntu Allwinner A10/A20, Allwinner H3, Amlogic, Freescale i.MX6x, Other boards with sticky stuff Or anything to have an overview for reports/status? just an idea.

in case it can help, I have copied the kernels I compile for my L-R1 + a package with older swconfig and/or zip with binary http://guillaume.romagny.free.fr/armbian_kernel/

@zador.b.s your latest updates for XU4/next (uboot 2016.09) are working fine on SD, I will check when possible for legacy edit: for eMMC, the sd_fusing.sh script is needed to wipe out uboot 2012.07 from "boot0" with 2016.09. if not wiped, booting is weird/KO.

my TODO list: 1/ (from zador.b.s) can you please test default branch on XU4 and report here? Last time command line didn't pass from u-boot to the kernel for some reason. 2/ looking why u-boot 2016.09 was not compiling. 3/ checking if 2016.05 can boot legacy kernel 4/ adding some documentation 5/ L-R1 https://github.com/igorpecovnik/lib/issues/511#issuecomment-256111716

Please provide your /etc/fstab file. Are you sure the network connexion is properly established when the automounting is started.

@zador.blood.stain 1/ ok, pulled 3 files changes. Your UUID architecture is like a charm. On the 3rd testing, the SDcard was booting & system running (Jessie). I guess eMMC will be working as well because it was working with root=LABEL=rootfs (to be confirmed yet) 2/ initrd was already implemented and working for XU4 next, so no changes here. 3/ I will be looking when possible about why u-boot 2016.09 was not compiling. Still uboot 2016.05 can be used for both SD/eMMC (huge move from 2012.07) 4/ I will check if 2016.05 can boot legacy kernel 4/ I will be looking to add some lines of documentation boot details:

just a crazy idea A downgrade from jessie to wheezy https://duckduckgo.com/?q=downgrade+jessie+wheezy&ia=web

@dvachon please don't be so hard. Currently the small Armbian team is doing a lot, so much that I just cannot follow. There is a lot of progress and so far quite properly architectured. The only big issue, the team is too small and barely can communicate properly. That has led to a huge misunderstanding with zador.blood.stain & I (that I am sorry) What are your questions?

@jeansburger As Igor writes, last time I have tested jessie (& a little bit xenial). Wheezy/XU4 has been removed from the distro list for a while. [ o.k. ] Writing U-boot bootloader [ /dev/loop0 ] [ o.k. ] Signing and compressing [ Please wait! ] [ o.k. ] Done building [ /home/gr/gro_armbian/output/images/Armbian_5.21-GRO_Odroidxu4_Debian_jessie_4.8.4.7z [194M] ] [ o.k. ] Runtime [ 39 min ]

please, is it useful if I would test MBR/PARTUUID for XU4 (as Zador is thinking of)? (instead of the former LABEL testing I did for SD/eMMC for next) in case I am on the proper path as of http://forum.armbian.com/index.php/topic/2454-fel-mass-storage-or-writing-images-directly-to-emmc/?p=18115

Hello TK, were the SD & eMMC using GPT or MBR? I am a little lost about PARTUUID which is/was originally dedicated to GPT OR is there now an option (clean "trick") to identify datas (partition) in the MBR as a PARTUUID? Zador wasn't clear about this point. Only mentioning about shorter version of PARTUUID (edit) is it http://www.gossamer-threads.com/lists/linux/kernel/1583540 ? When reading eMMC spec - boot0/1 are SLC NAND memory (Odroid size = 4MB) - rpmb is a security feature mimicking a partition (4MB) - the remaining of the eMMC is MLC NAND for Samsung (if not TLC or TLC3D for other brands) the eMMC provided by Odroid was sold with uboot 12.07 on the boot0 partition. boot1 is empty. uboot 12.07 was switching to mmcblk0 to fetch the initrd / kernel edit: uboot 12.07 can be replaced by uboot 16.05 (on boot0) and the XU4 is working fine with a single partition. eMMC is a standard so I guess the cards will load uboot from boot0 and fetch the kernel on the mmcblk0/1 partition so: SD card install: can only use MBR (to keep space for uboot) eMMC: can use either MBR or GPT as uboot/blobs are on boot0 partion edit 2: https://wiki.gentoo.org/wiki/GRUB http://lxr.linux.no/#linux+v3.8/init/do_mounts.c#L190

Hello Igor, I am not sure it is even possible in general. At least, the XU4 eMMC, the PC-SD2eMMC converter only displays the /dev/mmcblk0 (& not the other partitions boot0 boot1 rpmb) And what I have tested so far, you need a kernel with eMMC magement to access all the partitions.

@nz1 Armbian crew advises to buy legitimate Samsung EVO sdcard as reliable (32GB / 8 USD at gearbest) your story shows the card (or writer) is very low quality but still working when writing @ 1MB/s (which is lesser than "class 2")